The invention generally relates to a method of making a semiconductor device, and more particularly, to a method of making semiconductor pillar structures.
Devices made from semiconductor materials are used to create memory circuits in electrical components and systems. Memory circuits are the backbone of such devices as data and instruction sets are stored therein. Maximizing the number of memory elements per unit area on such circuits minimizes their cost and thus is a primary motivation in the designing of such circuits.
As the dimensions for structures formed on a semiconductor wafer diminish, tools currently available to create these devices reach their limits. By way of example, currently available 193 nanometer immersion tools will fail to create structures with a pitch of less than about 80 nm (i.e., with a half pitch of less than about 40 nm). To fabricate features smaller than this with the currently available tools, one must use more complicated processes. One such process is the technique of double exposure/double patterning. Another is the use of sidewall spacers, formed on a template pattern which is then removed. The sidewall spacers are then used as mask during etching of the underlying film or films.
For simple, one-dimensional, regular line-and-space patterns, both of these techniques have the effect of dividing the photolithographically-produced pitch by two. In this way, the resolution capability of a given photolithography tool can be extended.
However, for a two-dimensional pattern of regularly-spaced pillars, the double-patterning scheme extends the pitch by a factor of the square root of 2. The sidewall spacer method, as-is, cannot be used at all since such a scheme would produce regularly spaced cylindrical annuli, rather than solid pillars.